We should be more ambitious with the software we build
Points and comments are a snapshot, not live.
Dan Luu argues that AI tools now make ambitious software projects feasible.
Michael Malis's pgrust project rewrites Postgres in Rust for performance, uncovering real Postgres bugs. Dan Luu built a regex engine, a faster ripgrep fork, a Rust interpreter, and a superhuman Azul AI in hours with AI help. He says developers should pursue projects that were previously too large, though Yossi Kreinin warns that most people lack the ability to constrain AI-generated code, leading to overfitted, unmaintainable software.
What commenters are saying
The thread splits into two camps. One side celebrates AI as a creative accelerator: "I was a fantastic IC and built really amazing software. I'm an even better AI-empowered engineer." The other side fears job loss and commoditization: "The AI can do for $100 what took you 3 months full time" and "what differentiates software now will no longer be the care put into it but marketing and trends." Some argue that business-minded builders can tailor software to niche customers, while others call for a cultural shift away from racing to the bottom.